RIP, Anki: Yet Another Domestic Robotics Company

Today brings sad news in the world of consumer robotics. Anki, maker of Vector, a toy-like countertop robot, is closing its doors and hundreds of people are losing their jobs. This is a brutal answer to a question asked last summer: of course, domestic robots like Vector are cute, but is it really the will of consumers? Apparently, it's not enough customers for Anki, but here's what is much less clear: beyond the success of Roomba, what do we want on earth sophisticated machines at home?

Anki was not near the first company to try to populate the house of social robots – people have been doing it since the 80s. Almost all, however, failed – two important ones last year. There was Jibo, the stationary desktop robot who danced and died an untimely death, and Kuri, the R2-D2 machine who turned to take pictures of your dinner before embarking on the Grand Concert robotics in the sky.

The problem with both, and now with Anki's Vector, is that none of them really did a lot. Unlike Roomba and Rosie, mythically capable of Jetsonthey do not have any particular work to do. They were all cute, certainly, and undeniably charismatic, but three are the subject of a curse: Vector, Jibo and Kuri show that it is impossible to sell a domestic robot on its own account. The machine can not just be a companion – it must be a worker.

It may not be harder to do this than in the chaos of the house, what the robotics call an "unstructured" environment. For decades, robots have run the factory because it is a structured environment that is easy to master. They can take the prescribed routes and perform very repetitive maneuvers by facing few or no obstacles, unexpected. At home, robots have to face stairs and children and become cluttered on the ground. And they have to do it without being able to manipulate objects, perhaps the biggest hanging problem of robotics. Nobody gave a pair of hands to a domestic robot because he would be unable to grab objects.

Vector, Jibo and Kuri have made enormous progress, it is to navigate the complexities of human-robot interaction. "Having a built-in character allows us to do things that are not socially very acceptable from any device," said Hanns Wolfram Tappeiner, co-founder and president of Anki, during the announcement of Vector . "Vector does not sit there and wait until you ask him something." He was shaking and swaying around the counter. Unlike a home appliance, the machine was fun – a robot as we expected.

But in the end, Vector could not muster the utility of a typical device. The popularity of assistants like Amazon's Alexa did not make the company's job easier. Like Roomba and his sweep, these assistants do very well one thing. And they do this by sitting on a counter, which means they do not have to navigate the chaos of the house. If domestic robots can not yet take you a beer in the refrigerator, bend clothes or pick up an object literally, is their mobility of great value? And if they are not as smart as Alexa, is it useful to befriend them?

Domestic robot manufacturers are also struggling to keep up with the ability of Amazon and Google to intervene quickly on their voice assistants. A material like Vector's is a totally different story. If the company designs a robot and puts it on the market without success, it may take years to redefine it.

The ultimate goal for domestic robots may not even be to build a new best friend, but to make robots disappear in the house. "In the case of Roomba, it is only a vacuum cleaner, we do not think much like a robot," says Cory Kidd, developer of Mabu, a medical robot that interacts with the patients. "And I think it's really a success, when people think about these things in terms of functionality and the problem they're trying to solve." Think of the microprocessor: if you wear an old digital watch like me, you will not do it. Think of it as a computer because it has become more of a part of you than something unique. But it's really a computer.

Yet, what domestic robots have to do, is this unpublished charm. This is where the human-robot interaction bit becomes really interesting. Watching Kuri and Vector rolling beep and following your orders was a fascinating thing to see. Essentially, their physical presence makes it possible to create a link with a machine loaded with much more charisma than what a fixed voice assistant can muster. This is the promise of domestic robots: the agency illusion may not make them particularly useful at the moment, but they are forging a new and strange relationship between people and machines. Robots are no longer just tools: they are friends. And maybe someday (maybe once they've sprouted their arms and hands), they'll be so ingrained at home, like Roomba, that we'll forget they're robots.

Vector may have failed to gain enough hearts to survive, but Anki has made invaluable progress in clarifying what this new man-machine relationship might look like. "They were at the forefront of social robot design and have invested a lot to get a good interaction," says Kate Darling, a robot scientist at MIT, who studies human-robot interaction. "It was a wonderful product, but they had a hard time developing it – toy to something else."

Goodbye, sweet vector. May you find a use in the Gig Gig in the Sky.

More great cable stories